[nattonrice]

It is much sexier than a breadboard!

 

[johnwmclean]

...and you will need one for your EHHA Rev A.

 

[Nebby]

Next step would be to add a heater element to allow testing at operating conditions
 
Very nice build, I'll have to do something similar.

 

[jdkJake]

...and you will need one for your EHHA Rev A.

Already have two sets matched up just for that purpose.

 

[jdkJake]

Next step would be to add a heater element to allow testing at operating conditions


 


Very nice build, I'll have to do something similar.

Thanks. Thermocouple and a heat gun would probably work just fine if you really wanted to take it over the top!

 

[Nebby]

But if you're shooting for over the top, a heatgun won't keep the device under test at a consistent temperature

 

[jdkJake]

Point taken.

So I guess a full-up, temperature controlled, directed heating element is in order. Preferably under microprocessor control.

 

[wink]

Nah. Just overheat the dude until it's a pile of carbon, and it all becomes academic.

 

[nikongod]

You could just let the mosfet heat up naturally to a reasonable temperature.
 
on a slightly related note:
If you are seriously serious about this absurdly absurd absurdity
Build a little cover/enclosure for the test device. In my last FET matching efforts (to92 parts, but whatever) the FETs would drift when breathed on hard. 

 

[luvdunhill]

yeah, that's not going to work very well. You need to match the devices at their target operating junction temperature and at the target current and voltage. I used a ceramic heater control block of aluminum to do the job. The heating element is in the back and you can see the probe:
 

 
Also, I'd recommend an extra supply to bias the devices, in this case I used a battery with a simple adjustable regulator to vary the bias. This way I could control the exact current flowing through the device and matched this to the target current of the circuit, in this case it was 2.000A (the meter on the left, and see the bench supply reading as well).
 
To give you an idea how the device behaves as it warms up, here's some data taken off the Fluke meter in the picture:
 

 
Once you get things stable from a voltage, current and temperature standpoint, your matches will be very accurate. Here's a run for one device after it was dialed in:
 

 
That's a variance of 0.0004V over 2.5 minutes, which in this case was how long I tested each device to get a stable reading. I then saved the data for each device and used that to correlate matches via a Python script which spit out the "best matches" based on plot data.

 

[luvdunhill]

Quote:

If you are seriously serious about this absurdly absurd absurdity
Build a little cover/enclosure for the test device. In my last FET matching efforts (to92 parts, but whatever) the FETs would drift when breathed on hard. 

like this?
 

 
...been there done that
 

 

[luvdunhill]

oh, before someone says something about the use of a crappy meter that yellow meter isn't a yellow meter...it calculates hfe directly from wires that go to that box...

 

[ujamerstand]

Jesus. What devices are these?

 

[luvdunhill]

The big ones are Toshiba 2SJ201/2SK1530. Small ones are 2SA970/2SC2240. Medium ones are 2SA1145/2SC2705.

 

[Nebby]

Your insane matching quest was what came to mind as soon as I saw the matching device.